You analyze each conversation with an LLM: summarize it, add tags, identify problematic tools, etc. The metrics go to management, some docs are auto-generated and added to the company knowledge base like all other company docs.
It’s like what they do in support or sales. They have conversational data and they use it to improve processes. Now it’s possible with code without any sort of proactive inquiry from chatbots.
You analyze each conversation with an LLM: summarize it, add tags, identify problematic tools, etc. The metrics go to management, some docs are auto-generated and added to the company knowledge base like all other company docs.
It’s like what they do in support or sales. They have conversational data and they use it to improve processes. Now it’s possible with code without any sort of proactive inquiry from chatbots.